Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Peng Wang has a total of 30 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Can Large Language Models Resolve Semantic Discrepancy in Self-Destructive Subcultures? Evidence from Jirai Kei

    by Peng Wang, Xilin Tao, Siyi Yao, Jiageng Wu, Yuntao Zou, Zhuotao Tian, Libo Qin, Dagang Li

    Self-destructive behaviors are linked to complex psychological states and can be challenging to diagnose. These behaviors may be even harder to identify within subcultural groups due to their unique expressions. As large language models (LLMs) are applied across various fields, some researchers have begun exploring their application for detecting self-destructive behaviors. Motivated by this, we investigate self-destructive behavior detection within subcultures using current LLM-based methods. However, these methods have two main challenges: (1) Knowledge Lag: Subcultural slang evolves rapidly, faster than LLMs' training cycles; and (2) Semantic Misalignment: it is challenging to grasp the specific and nuanced expressions unique to subcultures. To address these issues, we proposed Subcultural Alignment Solver (SAS), a multi-agent framework that incorporates automatic retrieval and subculture alignment, significantly enhancing the performance of LLMs in detecting self-destructive behavior. Our experimental results show that SAS outperforms the current advanced multi-agent framework OWL. Notably, it competes well with fine-tuned LLMs. We hope that SAS will advance the field of self-destructive behavior detection in subcultural contexts and serve as a valuable resource for future researchers.

  • Top Cited Papers
    Open Access

    S 3 Agent: Unlocking the Power of VLLM for Zero-Shot Multi-Modal Sarcasm Detection

    by Peng Wang, Yongheng Zhang, Hao Fei, Qiguang Chen, Yukai Wang, Jiasheng Si, Wenpeng Lu, Min Li, Libo Qin

    Multi-modal sarcasm detection involves determining whether a given multi-modal input conveys sarcastic intent by analyzing the underlying sentiment. Recently, vision large language models have shown remarkable success on various of multi-modal tasks. Inspired by this, we systematically investigate the impact of vision large language models in zero-shot multi-modal sarcasm detection task. Furthermore, to capture different perspectives of sarcastic expressions, we propose a multi-view agent framework, S 3 Agent, designed to enhance zero-shot multi-modal sarcasm detection by leveraging three critical perspectives: superficial expression , semantic information , and sentiment expression . Our experiments on the MMSD2.0 dataset, which involves six models and four prompting strategies, demonstrate that our approach achieves state-of-the-art performance. Our method achieves an average improvement of 13.2% in accuracy. Moreover, we evaluate our method on the text-only sarcasm detection task, where it also surpasses baseline approaches.


  • Open Access

    Improving Consistency Identification in Task-oriented Dialogue Through Multi-Agent Collaboration

    by Peng Wang, Shuo Li, Ruoxi Zhou, Qiguang Chen, Xiao Xu, Hao Fei, Dagang Li, Wanxiang Che, Libo Qin

    Consistency identification in task-oriented dialog (CI-ToD) typically consists of three sub-tasks: User Query Inconsistency (QI) identification, Dialogue History Inconsistency (HI) identification, and Knowledge Base Inconsistency (KBI) identification, which aim to determine inconsistent relationships between system response and user query, dialogue history, and knowledge base. Previous approaches focus on the exploration of deep learning models for CI-ToD. While these models achieve remarkable progress, they still rely on large amounts of labeled data, which is hard to achieve in real-world scenarios. Motivated by this, in the paper, we aim to explore large language models for CI-ToD, which do not require any training data. In addition, we further introduce a multi-agent collaboration framework (MAC-CIToD) to model the interaction across three sub-tasks in CI-ToD, including (1) Full Connection paradigm, (2) Cycle Connection paradigm, and (3) Central Connection paradigm, which effectively builds interaction across QI, HI, and KBI. Experiments on the standard benchmark reveal that our framework achieves superior performance. Additionally, we compare MAC-CIToD with the most advanced trained approaches and find that its zero-shot performance on most metrics even surpasses that of models after training on the CI-ToD dataset.


  • Open Access

    X-WebAgentBench: A Multilingual Interactive Web Benchmark for Evaluating Global Agentic System

    by Peng Wang, Ruihan Tao, Qiguang Chen, Mengkang Hu, Libo Qin

    Recently, large language model (LLM)-based agents have achieved significant success in interactive environments, attracting significant academic and industrial attention. Despite these advancements, current research predominantly focuses on English scenarios. In reality, there are over 7,000 languages worldwide, all of which demand access to comparable agentic services. Nevertheless, the development of language agents remains inadequate for meeting the diverse requirements of multilingual agentic applications. To fill this gap, we introduce X-WebAgentBench, a novel multilingual agent benchmark in an interactive web environment, which evaluates the planning and interaction performance of language agents across multiple languages, thereby contributing to the advancement of global agent intelligence. Additionally, we assess the performance of various LLMs and cross-lingual alignment methods, examining their effectiveness in enhancing agents. Our findings reveal that even advanced models like GPT-4o, when combined with cross-lingual techniques, fail to achieve satisfactory results. We hope that X-WebAgentBench can serve as a valuable benchmark for multilingual agent scenario in real-world applications.

  • Open Access

    X-WebAgentBench: A Multilingual Interactive Web Benchmark for Evaluating Global Agentic System

    by Peng Wang, Ruihan Tao, Qiguang Chen, Mengkang Hu, Libo Qin

    Peng Wang, Ruihan Tao, Qiguang Chen, Mengkang Hu, Libo Qin. Findings of the Association for Computational Linguistics: ACL 2025. 2025.


  • Open Access

    TRSRD: a database for research on risky substances in tea using natural language processing and knowledge graph-based techniques

    by Yongmei Wang, Peng Wang, Yongheng Zhang, Siyi Yao, Zhipeng Xu, Youhua Zhang

    During the production and processing of tea, harmful substances are often introduced. However, they have never been systematically integrated, and it is impossible to understand the harmful substances that may be introduced during tea production and their related relationships when searching for papers. To address these issues, a database on tea risk substances and their research relationships was constructed. These data were correlated by knowledge mapping techniques, and a Neo4j graph database centered on tea risk substance research was constructed, containing 4189 nodes and 9400 correlations (e.g. research category-PMID, risk substance category-PMID, and risk substance-PMID). This is the first knowledge-based graph database that is specifically designed for integrating and analyzing risk substances in tea and related research, containing nine main types of tea risk substances (including a comprehensive discussion of inclusion pollutants, heavy metals, pesticides, environmental pollutants, mycotoxins, microorganisms, radioactive isotopes, plant growth regulators, and others) and six types of tea research papers (including reviews, safety evaluations/risk assessments, prevention and control measures, detection methods, residual/pollution situations, and data analysis/data measurement). It is an essential reference for exploring the causes of the formation of risk substances in tea and the safety standards of tea in the future. Database URLhttp://trsrd.wpengxs.cn


  • Open Access

    Building Natural Language Interfaces Using Natural Language Understanding and Generation: A Case Study on Human–Machine Interaction in Agriculture

    by Yongheng Zhang, Siyi Yao, Peng Wang, Hao Wu, Zhipeng Xu, Yongmei Wang, Youhua Zhang

    The human–machine interaction of existing agricultural measurement and control platforms lacks user-friendliness and requires manual operation by trained professionals. The recent development of natural language processing technology may bring some interesting changes. We propose a pipeline for building a natural language human–machine interaction interface to provide a better interaction for agricultural measurement and control platforms. Our construction process uses a new method of collecting training data based on the dynamic tuple language framework to synthesize natural language commands entered by the user into structured AOM statements (Action-Object-Member). To construct a mapping of the human–machine interface from natural language commands to AOM invocations, we propose an end-to-end framework that uses a special mask mechanism to improve the BERT-based Seq2Seq model to capture global sequence relations. Experimental results of data collection methods and NL2AOM demonstrate that our pipeline has good performance and a reasonable response time. Finally, we developed desktop and mobile platform applications based on the proposed model and used them in real agricultural scenarios.


  • Open Access

    Long-Term Water Quality Prediction Using Integrated Water Quality Indices and Advanced Deep Learning Models: A Case Study of Chaohu Lake, China, 2019–2022

    by Siyi Yao, Yongheng Zhang, Peng Wang, Zhipeng Xu, Yongmei Wang, Youhua Zhang

    The rapid development of urban industrialization has had many negative effects on the quality of water sources around cities. Long-term prediction of water quality can be of great help to the conservation of water environment. This case tries to use several popular deep learning models, such as RNN, LSTM, MLP, and Transformer-based models to predict the long-term integrated water quality index in the Chaohu Lake area. The dataset is derived from daily monitoring data from four monitoring sites within Chaohu Lake from 2019 to 2022, and the long-term prediction performance of the model is evaluated using MAE and MSE as evaluation metrics. The experimental results showed that all models selected in this case achieved good results within the study area, but Informer performed more prominently (MSE = 0.2455, MAE = 0.2449) as the length of the prediction series increased. Our results demonstrate the effectiveness of popular deep learning models in the field of WQI prediction, especially the significant advantage of transformer-based models represented by Informer in long-term water quality prediction, which will further provide an effective modern tool for water quality monitoring and management.